Transformative role of machine learning in design optimization of reinforced concrete frames

Abstract

Design optimization in reinforced concrete (RC) frames is crucial for achieving more efficient, cost-effective, and safe constructions. This involves determining optimal section sizes and reinforcing schemes to enhance structural performance, reduce construction expenses, and conserve natural resources. An optimized design contributes to streamlined construction procedures, shorter construction times, and improved overall sustainability. This paper critically evaluates the current literature on machine learning (ML) applications in the design optimization of RC structures, focusing on the period from 1997 to 2023. The study employs a comprehensive methodology, including bibliometric analysis, to analyze trends, research collaborations, and publication patterns. The results highlight the increasing interest in ML for RC frame optimization. Key applications of ML in design optimization include material characterization, design exploration, optimization algorithms, sensitivity analysis, predictive modeling, structural health monitoring, design code compliance, uncertainty quantification, data-driven design decisions, and design collaboration. The study identifies significant ML algorithms used in optimization, such as the trial-and-error method, linear programming (LP), and nonlinear programming (NLP). Overall, the paper provides insights into the evolving landscape of ML applications in RC frame design, emphasizing the potential for interdisciplinary collaboration and future research directions.